Srikanth Parameswaran

Social Support in Online Health Communities: A Social-Network Approach

We develop a social network based model of user-generated content (social support) in online health communities. Grounded in the social presence and the social network theories, we hypothesize the differential impacts of an online health community members brokerage and influence in the web-of-support on a) extent of social support, and two self-absorption linguistic features namely b) self-focused and c) socially-focused words in social support text. Variables were operationalized using text mining and social network analyses. We empirically tested our model using panel-data collected from an online health community for diabetics. Higher brokerage results in more support from the member; however, the marginal effect is decreasing. Higher influence results in reduced support; however, the effect is increasing at higher levels. Higher brokerage and influence results in more socially-focused support; however, the marginal effect is decreasing. Higher brokerage and influence results in less self-focused support; however, the marginal effect is increasing.

Can online social support be detrimental in stigmatized chronic diseases? A quadratic model of the effects of informational and emotional support on self-care behavior of HIV patients

Objective We studied the impact of online social support on patient self-care behavior in an online health community for human immunodeficiency virus (HIV) patients. We conceptualized emotional and informational support provided by community members into nuanced sub-dimensions. We explored how the direct and interaction effects of these sub-dimensions impact the self-care behavior of a support seeker. Methods We used data from 330 255 posts in 30 050 threads from POZ, an online health community for HIV patients. Our key variables-self-care behaviori, objective informationj, experiential supportj, and emotional tonej-were operationalized using linguistic analysis with self-generated dictionaries and Python libraries. We tested our hypotheses using Tobit regression. Results Out of 6 null hypotheses, 5 were rejected. Objective information and emotional tone had an inverted-U relationship with self-care behavior. Experiential information and community involvement were positively related to self-care behavior. Community involvement amplified the inverted-U relationship between emotional tone and self-care behavior. No significant interaction effect was found between experiential support and objective information. Conclusions Beyond a threshold, both informational and emotional online social support had a deleterious impact on self-care behavior of HIV patients. Our results suggested that caution should be exercised in the use of online health community interventions for HIV patients, and perhaps patients with other stigmatized chronic diseases.

Use of Gamified Social Media with Home Telemonitoring for Patient Self-Management in Poorly Controlled Medicaid Diabetics: A Pilot Study of Health Outcomes, Social Influences, and Habit Formation

This paper summarizes a poster which presents an overview of a proposed healthcare study. The study is aimed to motivate patients, enhance patient self-care of diabetes and improve health outcomes. This study proposes to evaluate the effectiveness of a gamified social media app with a competitive game element in conjunction with home remote patient monitoring (the proposed technology intervention) in a sample of Medicaid diabetic patients with poorly controlled diabetes.